Method

How Iris values a point

Every dollar figure in Iris rests on one number per programme: what a single point is worth, in cents. This page is where that number comes from, who published it, and what it cannot tell you.

The rule

The value of a point is the median of three published valuations, each with a named author and a date, rounded down to the nearest hundredth of a cent. For every programme we read The Points Guy, One Mile at a Time and Frequent Miler. We take the middle of the three. Where one of them does not cover a programme, Upgraded Points takes its place, and the table below marks it. Where fewer than three publish a programme, Iris carries no dollar value for it at all.

Why a median and not an average

On the same currency, the three sources disagree by as much as 0.65 of a cent. An average pulls that disagreement into the result. A median discards both ends and keeps the middle opinion, which is the point: we are not trying to be right about the extremes, we are trying not to be captured by one of them.

Why we round down, and why to a hundredth

Down, because Iris exists to tell you when your points are losing value. Understating what you hold is the cheap error; overstating it makes every warning louder than the facts.

To a hundredth rather than a twentieth, because the rounding grid decides how much information gets thrown away. On a 1.50-cent point, five hundredths of a cent is noise. On a 0.59-cent hotel point, it is a 7% cut that none of the sources supports. The grid has to survive the cheapest currency on the list, not the most expensive.

Who these publications are

The sources behind every figure

Read them directly: The Points Guy monthly valuations and their stated methodology · One Mile at a Time valuation guide · Frequent Miler Reasonable Redemption Values · Upgraded Points valuations.

What we are not going to hide from you

All of these publications earn commission on credit card applications. Iris does not. Three sources reduce variance; they do not remove a bias the three of them share. And the bias has a known direction: a high point value makes a credit card look better, so the pressure is upward. Since what Iris gives you is a warning about loss, an inflated valuation would overstate every loss Iris reports. That is why Iris takes the middle figure and rounds it down.

One of them discloses, in an editor's note, that its founder is an adviser to and an investor in one of the programmes it values. That programme also carries the highest figure in their table and the widest gap from the other sources for the same currency. This is not an insinuation — it is written on their own page, and it is exactly the case a median exists to absorb.

The sources are not fully independent at the measurement layer. For airline and hotel programmes, two of them draw on the same underlying award-search data. For the five bank currencies they are methodologically independent. So "three independent sources" is true for bank points and weaker for the rest, and you should read the table knowing that.

Iris does not measure award prices itself. Until it does, the number you see is a median of public opinion, attributed here rather than presented as our own finding. If Iris later publishes a measured value, it will carry its own name and its own method. It will not silently replace this one.

Avios is one currency

British Airways Club, Iberia Plus and Qatar Privilege Club transfer between one another at 1:1, so they carry a single value. Where a source publishes a single figure for Avios rather than one per programme, we apply it to all three. Publishing three different numbers would be Iris asserting an arbitrage that any holder undoes in two clicks.

When we don't know, we don't show a number

If any programme in your selection has no dated valuation on file, Iris shows your balance in points and no dollar figure at all. A number without a source behind it is worse than no number, because it looks like knowledge.

Programme valuations

UP marks a programme where Upgraded Points stood in for Frequent Miler.

Changes to this method

September 15, 2026 — v1. Median of three published valuations, rounded down to the nearest hundredth of a cent.